The background description provided herein is for the purpose of generally presenting the context of the disclosure. Work of the presently named inventor(s), to the extent the work is described in this background section, as well as aspects of the description that may not otherwise qualify as prior art at the time of filing, are neither expressly nor impliedly admitted as prior art against the present disclosure.
Computers in a network can be arranged into many different topologies. For example, a network can be configured in a star topology, a mesh topology, a ring topology, and so on. In a network configured in a star topology, a central point (e.g., an access point (AP)) communicates directly with client devices and controls network functions (e.g., key generation/distribution, network access) for the client devices. By contrast, in a network configured in a mesh topology, each device communicates with many other devices to create a network that resembles, for example, a web of connections. A mesh network employs distributed techniques that do not rely on a central point to control network functions. Employing distributed techniques can be computationally intensive. For example, on the order of N2 communications (where N is the number of devices in the network) can be necessary to perform functions in a mesh network.
As an example, consider performing a key distribution function in a star topology network versus a mesh topology network. In a star topology network, only a single group key is required. This is because all messages are sent directly from the access point to devices in the network and devices do not communicate with each other directly. Thus, distributing a new key to all devices in the star topology network requires, for example, 4N messages.
However, in a mesh topology network, each device uses a specific group key when communicating with another device. This is because devices can communicate directly without using a central point. Accordingly, each device needs to know the specific group keys of all of the devices in the mesh topology network. The group keys for each device must be exchanged between every other device on the network, which is performed by exchanging multiple messages between each pair of devices. To distribute keys for each device in a network of N devices, 4N(N−1) communications may be required. Now, consider a network with 30 devices (N−30). A star topology network uses 120 communications (4×30) to distribute a new key. By contrast, a mesh topology network uses 3,480 communications (4×30(30−1)) to distribute new keys. In this example, the mesh topology network requires twenty nine times more communications than the star topology network. Accordingly, distributing keys in the mesh topology network uses many communications that can cause difficulties with efficiently establishing shared keys.